遺傳算法參數(shù)自適應(yīng)控制及收斂性研究.pdf

1、中南大學(xué)碩士學(xué)位論文遺傳算法參數(shù)自適應(yīng)控制及收斂性研究姓名：鄺溯瓊申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：電工理論與新技術(shù)指導(dǎo)教師：喻壽益20090521ABSTRACTTheperformanceofgeneticalgorithm(GA)isobviouslyenslavedtothecrossoverprobabilityPcandmutationprobabilityP朋ForthevaluesofpcandP所areinvariable，th

2、ereareproblemsasthelowconvergencerateandprematurephenomenoninsimplegeneticalgorithmThoughadaptivegeneticalgorithm(AGA)hasbeenproventobeabletoadjustthecontrolparametersaccordingtothefitnessoftheoverallpopulation，itisnotsu

3、itablefordifferentproblemsfortherulestoadjustpcandp，行mustbevariedwithoptimizationquestionsSOastofindtheoptimalsolutionfasterandmorepreciselyItisofgreatimportancetoadjusttheparametersadaptivelyConvergencetheoryisacoretheo

4、reticalissueongeneticalgorithmsThesystematismanduniversalismofthetheoriesarenotenoughSo，studyingtheconvergencetheoryofgeneticalgorithmsistheinevitablerequirementforthetheoryandapplicationofGATherefore，parameteradaptiveco

5、ntrollingandconvergencetheoryforgeneticalgorithmsarestudiedinthispaperAkindofnestedfuzzyadaptivegeneticalgorithm(NFAGA)whichissuitablefordifferentoptimizationproblemsisrealized，thesimulationplatformofthenewalgorithmisimp

6、lemented，andthesimulationexperimentsinfunctionoptimizationproblemswithseveralothergeneticalgorithmsarecarriedoutFortheconvergencetheorytheconvergenceandconvergencerateofelitistgeneticalgorithm(EGA)basedonmartingaleapproa

7、chisstudiedThedetailedcontentsandinnovativeworkofthispaperareasfollows：1Accordingtothedefectsforregulationlawoffuzzyadaptivegeneticalgorithmbasedonexpertise(EFAGA)，theselfleamingmechanismtoupdateandoptimizetheknowledgeba

8、sisofthefuzzycontrollerisstudiedOntheonehand，thevaluesofpcandp，打areadjustedaccordingtotheevolutionalprocessbyusingthefuzzycontrolmethod；ontheotherhand，thefuzzyrulesandtheparametersofthemembershipfunctionsareoptimizedbyan